R/Sampler.R

In paradox: Define and Work with Parameter Spaces for Complex Algorithms

#' @title Sampler Class
#'
#' @description
#' This is the abstract base class for sampling objects like [Sampler1D], [SamplerHierarchical] or [SamplerJointIndep].
#'
#' @template param_param_set
#'
#' @family Sampler
#' @export
Sampler = R6Class("Sampler",
  public = list(
    #' @field param_set ([`ParamSet`])\cr
    #' Domain / support of the distribution we want to sample from.
    param_set = NULL,

    #' @description
    #' Creates a new instance of this [R6][R6::R6Class] class.
    #'
    #' Note that this object is typically constructed via derived classes,
    #' e.g., [Sampler1D].
    #' @param param_set ([`ParamSet`])\cr
    #'   The [`ParamSet`] to associated with this `Sampler`.
    initialize = function(param_set) {
      assert_param_set(param_set, no_untyped = TRUE)
      self$param_set = param_set$clone(deep = TRUE)
    },

    #' @description
    #' Sample `n` values from the distribution.
    #'
    #' @param n (`integer(1)`).
    #' @return [`Design`].
    sample = function(n) {
      assert_count(n) # we do argcheck on toplevel
      Design$new(self$param_set, private$.sample(n), remove_dupl = FALSE) # user wants n points, dont remove
    },

    #' @description
    #' Helper for print outputs.
    #' @param ... (ignored).
    format = function(...) {
      sprintf("<%s>", class(self)[1L])
    },

    #' @description
    #' Printer.
    #'
    #' @param ... (ignored).
    print = function(...) {
      catf(format(self))
      catf("For params: %s", str_trunc(str_collapse(self$param_set$ids()), width = 40L))
      private$.print()
    }
  ),
  private = list(
    .sample = function(n) stop("abstract"), # inheriting classes have to implement this
    .print = function() {
    } # inheriting classes can overwrite to add lines
  )
)